Attempt to load model weights instead of direct model from file

app.py

@@ -1,5 +1,5 @@
 import tensorflow as tf
-from tensorflow import keras
+from tensorflow.keras import layers, models
 import gradio as gr
 import numpy as np
 import cv2
@@ -9,8 +9,48 @@ classes = ["Abyssinian", "Bengal", "Birman", "Bombay", "British Shorthair", "Egy
 example_images = ["examples/" + f for f in os.listdir("examples")]
 
 img_size = 400
-model = tf.keras.models.load_model("/home/user/app/CatClassifier.keras")
-print(os.getcwd())
+num_classes = 12
+
+# Create CNN model architecture and apply weights from file
+    def create_model():
+    model = models.Sequential()
+    model.add(layers.RandomFlip("horizontal_and_vertical"))
+    model.add(layers.RandomRotation(0.2))
+    model.add(layers.RandomZoom((0, 0.2)))
+    model.add(layers.Rescaling(1./255))
+    
+    model.add(layers.Conv2D(8, 3, activation="relu"))
+    model.add(layers.BatchNormalization())
+    model.add(layers.MaxPooling2D())
+    
+    model.add(layers.Conv2D(16, 3, activation="relu"))
+    model.add(layers.BatchNormalization())
+    model.add(layers.MaxPooling2D())
+    
+    model.add(layers.Conv2D(32, 3, activation="relu"))
+    model.add(layers.BatchNormalization())
+    model.add(layers.MaxPooling2D())
+    
+    model.add(layers.Conv2D(64, 3, activation="relu"))
+    model.add(layers.BatchNormalization())
+    model.add(layers.MaxPooling2D())
+    
+    model.add(layers.Conv2D(92, 3, activation="relu"))
+    model.add(layers.BatchNormalization())
+    model.add(layers.MaxPooling2D())
+    
+    model.add(layers.BatchNormalization())
+    
+    model.add(layers.Flatten())
+    
+    model.add(layers.Dense(1024, activation="relu"))
+    model.add(layers.Dropout(0.5))
+    model.add(layers.Dense(512, activation="relu"))
+    model.add(layers.Dense(num_classes, activation="softmax"))
+    model.load_weights("CatClassifierWeights.h5")
+    return model
+
+model = create_model()
 
 def model_predict(image):
     image = cv2.resize(image, (img_size, img_size))